A video codec's job is delivering acceptable quality in as few bits as possible. AV1 does that better than the codecs that preceded it — meaningfully lower bitrate for equivalent quality — and it achieves that by doing considerably more computation during encoding.
That computation is the whole tradeoff, and hardware support is what resolved it.
What changed in 2026
- Hardware encoding became widespread. Graphics hardware and mobile chips with AV1 encoders removed the speed penalty for real-time and consumer use.
- Streaming platform adoption broadened. Delivery in AV1 to supporting devices became standard at major platforms, driven by bandwidth savings at scale.
- Live streaming adoption grew. With hardware encoders available, live use became practical rather than theoretical.
- The successor discussion started. Work on the next-generation royalty-free codec advanced, with deployment years away.
Codec comparison
| Codec |
Compression efficiency |
Encode speed |
Licensing |
Device support |
| Older widely-supported codec |
Baseline |
Fast |
Licensed |
Universal |
| Mid-generation codec |
Better |
Moderate |
Licensed; complex |
Broad |
| AV1 |
Best of the widely deployed |
Slow in software, fast in hardware |
Royalty-free |
Broad and growing |
| Next-generation successor |
Better still |
Very slow currently |
Royalty-free |
Minimal |
The licensing column explains the adoption pattern more than the technical one does. The mid-generation licensed codec offered strong efficiency and a licensing structure complex enough that many companies avoided it. A royalty-free alternative with comparable efficiency removed that obstacle, which is why an industry consortium formed around it.
Where it fits
Delivery at scale. If you serve a large volume of video, lower bitrate for the same quality means directly lower bandwidth cost. That is the case that justified the industry investment, and the savings are real at volume.
Bandwidth-constrained viewers. Better quality at low bitrates improves the experience on poor connections more than it does on fast ones.
Live streaming with hardware encoding. Now practical, and the quality-per-bitrate advantage helps where upload bandwidth is limited.
Not for production. Editing workflows want fast decode, frame-accurate seeking, and minimal generational loss. Intermediate codecs designed for editing serve that far better than any delivery codec, AV1 included.
Not for archival masters. Archive in the highest-quality format you can afford. Delivery codecs discard information by design; archiving in one bakes that loss in permanently.
The encoding cost
Software AV1 encoding is slow enough to be genuinely inconvenient for large libraries. Encoding presets trade speed against efficiency, and the faster presets give up a substantial share of the compression advantage — which sometimes eliminates the reason to use it.
Hardware encoders are dramatically faster and typically less efficient than a slow software encode at the same bitrate. For real-time use that tradeoff is obviously correct. For batch encoding a library where quality per byte matters and time does not, software with a slower preset produces better results.
Check decode support for your audience. Decoding is much cheaper than encoding and support is broad on recent devices, and older hardware may fall back to software decode, which drains battery on mobile.
Common mistakes
- Using it for editing masters. Wrong tool; use an intermediate codec.
- Archiving in a delivery codec. Bakes in loss permanently.
- Fast software presets. Give up much of the efficiency advantage.
- Assuming universal decode support. Older devices may fall back to software.
- Comparing codecs at the same bitrate without checking quality. Efficiency claims need quality-matched comparison.
FAQ
How much bandwidth does it save?
Meaningfully less bitrate for equivalent quality versus older codecs, with the exact figure depending on content and encoder settings. Test on your own material.
Do I need new hardware?
For fast encoding, yes — a device with a hardware AV1 encoder. For decoding, most recent devices support it.
Is it better than the licensed mid-generation codec?
Comparable efficiency with no licensing complexity, which is why it won broader adoption.
What about the next generation?
In development with substantially higher encoding cost and years from deployment. Not a reason to defer AV1.
Where to go next
For the next-generation licensed alternative, read H.266 VVC explained. For capture and production setups, streaming setup guide.